Author: Volker Böhm
Date: 07:37:25 05/29/04
Hi,
I know it needs many lines of code and it is error-prone but I am currently
writing a static mate-in-one detection.
I´ll experiment a bit with this code with nullmoves. As I extend if nullmove
returns mate in one I whant the nullmove check to find most mate in one
independent of current window. Perhaps I can use this code to extend even if I
decide not to try nullmoves.
On the other hand it takes time to calulate mate-in-one thus I´ll only check
common mate situations.
Currently I am looking for the following mating moves:
1. Queen adjacent to king (diagonal, vertical, horizontal) but not from behind
("behind": row below white king, row above black king)
2. Rook adjacent to king (vertical, horizontal) but not from behind
3. Rook or Queen on same row or column as king, but not from behind
Thus I do not check:
1. mate by moving a queen or rook behind the king
2. mate by moving a bishop
3. mate by moving a knight
4. mate by moving a pawn
5. mate by uncovering a piece
6. mate by casteling
7. mate by promotion
8. mates that cannot be defended because the defending piece is pinned
In my opinion the latter 8 cases are to seldom compared to the 3 cases above.
They have to be found by "normal" search.
What do you think? Witch cases do you or would you add to a static mate-in-one
detection?
Volker
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.